euxhenh/grinch

A library for differential expression and gene set enrichment analysis based on hydra and pydantic.

grinch is a (under development) Python library for single-cell data analysis aimed at reducing boilerplate code through the use of workflow 'config' files. This encourages reproducibility, minimizes bugs, and allows the construction of complex pipelines with zero code written.

The library was originally written to prevent overcrowding project directories with a myriad jupter notebooks and scripts, and replace them with short and self-explanatory configs. grinch supports many steps of the single-cell analysis pipeline, from normalization and unsupervised learning, to gene differential expression and gene set enrichment analysis.

Architecture

grinch uses the AnnData data format to store data matrices and annotations. It relies on hydra for parsing yaml config files and pydantic for data validation.

Usage

Example configs can be found in the conf directory. More docs to come...